AI Readiness Check for Microsoft 365
Before you roll out AI, find out what it will have access to. This is how I check a tenant, and where Raidiness helps.
By Ziggy Itjoejaree · Updated
What is an AI Readiness Check?
An AI Readiness Check looks at whether your Microsoft 365 environment, your data and your organization are ready for Copilot or agents. What's in order, what needs work, and what you simply don't know yet.
I always start with the use case, not the tenant. Who is going to use AI? For which tasks? With which information? And who makes the decisions? Without those answers, technical findings are just a list.
I built Raidiness to do the data collection and the checks. Connecting the results to how your organization actually works is still a human job.
Why check before you roll out Copilot?
Microsoft 365 Copilot works within the permissions a user already has. The truth is, in most tenants people can reach more than they should. Before Copilot, nobody noticed. With Copilot, that information is one question away. Microsoft's Copilot security guidance explains how permissions, oversharing and risk are connected.
That's why I want the data and governance conversation before the licenses are assigned, not after.
My Microsoft 365 AI readiness checklist
1. Scope: what are you rolling out?
Copilot Chat, Microsoft 365 Copilot or agents? Who are the users, which tasks, which data sources, and who owns the rollout? A pilot group of 50 people needs a different check than the whole tenant.
2. Identity, access and licensing
Check users, groups, licenses and identity settings for the scenario you picked. Do the selected users and workloads meet the requirements? If something couldn't be read because of missing permissions, write that down. A missing permission shouldn't look like a clean result.
3. Data access and SharePoint governance
Look at sharing settings, site owners and where sensitive content lives. Permission sampling shows you where to take a closer look. It doesn't prove that every file in the tenant is set up correctly.
4. Information protection and audit
Review sensitivity labels, DLP, retention and audit settings in Purview. Then check whether they actually cover the users and information in scope. A DLP policy that exists isn't the same as a DLP policy that works.
5. Teams, agents and connectors
Include the workloads your rollout will touch, like Teams meeting policies and Power Platform environments and connector policies. For a full picture of agents or shadow AI, you'll need other tools and a manual review.
6. People and ownership
Who owns the AI policy, training, support and the success criteria for the pilot? The intake captures those decisions. For every finding, agree who picks it up, what the next step is and how you'll check that it's fixed.
How I use Raidiness in an assessment
- Set the scope. Pick the AI scenarios, users and workloads.
- Run Raidiness. Use approved access and fill in the intake.
- Review the findings. Keep what was measured apart from what couldn't be checked.
- Agree the next steps. Go through the findings with the owners and turn them into a plan.
- Run it again after changes. A fresh run shows what really changed. Re-exporting an old report doesn't.
What I want at the end is a clear list of findings, open questions and decisions for the pilot. Raidiness delivers the technical part. The organization decides the priorities and whether the rollout goes ahead.
Questions I often get
Is this a Microsoft 365 Copilot readiness assessment?
Yes, Copilot is the main use case. Scope it to the AI features you actually plan to use, including chat or agent scenarios. Do note that Raidiness is my own independent project. It's not a Microsoft certification or approval.
Can a scan prove we're ready for AI?
No. A scan can only judge what it can see. It can't tell you whether every permission is right, whether people understand the policy or whether a use case will pay off. For that you need people who know the organization.
Does Raidiness find all shadow AI?
No. It doesn't give you a full inventory of every AI tool or agent in use. If shadow AI is part of your scope, add other discovery tools and talk to the business.
Can I run the check myself?
Yes, if you have the technical knowledge and approved access. Start with my Raidiness installation guide. Want to talk through your scope or findings? Contact me.
Further reading
Questions about Raidiness?
Tell me what you want to roll out and where you're stuck, and I'll take a look.
Get in touch